|
|
Post by danoxon on Jun 23, 2023 14:37:20 GMT -8
peterberg had mentioned he was not happy with the percentage of burns in DSR's that go into thermal runaway and thus enter a dirty burn phase. If I understand correctly, this is positive feedback from the draft pulling in more primary air, causing more gas production, more heat from gas-afterburn and thus increasing draft. There is an application of Venturi nozzles as flow limiters known as “sonic nozzle", "Critical Flow Venturi” or “Critical Flow Nozzle” www.flowsystemsinc.com/sonic-nozzles-and-critical-flow-venturis/#:~:text=A%20%E2%80%9Csonic%20nozzle%E2%80%9D%20(also,or%20any%20flow%20measurement%20device. Since DSR's already have a slot for gas passage, perhaps this could be more rigorously designed as a CFV, and prevent excess draft from being able to pull more primary air past the CFV limit. The questions is, can the venturi throat be made small enough to achieve critical flow without causing other problems? 1. Startup problems before draft established 2. venturi exit flow too smooth for good turbulent burn without other physical features added post-venturi exit. 3. Can a CFV even operate at the low pressures in a passively driven stove? -Dan
|
|
|
|
Post by peterberg on Jun 25, 2023 0:29:20 GMT -8
Hi Dan, the DSR1 has been discontinued, just because of the tendency to overfueling. This particular problem has been solved in the DSR3, by adding a second restriction at the end of the combustion traject. It seems that this second restriction is functioning as a limiter for the maximum air velocity in the first one, i.e. the port. No thermal runaway anymore, no restriction at all upto a tuned value, after that the limiting function kicks in. I don't have a scientific explanation for this phenomenon, I can only reflect on what I have done and measured through the course of the development of this latest version. The development is published in this thread: donkey32.proboards.com/thread/3710/dsr1-vortex-aspects-dsr3At the moment there are more than two dozen heaters sporting this DSR3 core and they all work very well, as far as I can tell. This core is even more critical than the original batchrocket, one little deviation and the behaviour will change. Not for the best, I might add. |
|
|
|
Post by danoxon on Jun 26, 2023 22:25:52 GMT -8
I read the development thread you pointed to. Interesting! I wonder if the top chamber's flow-limiting action is Critical-Flow-venturi like behavior of the end-port. Has anyone puzzled out why the csa of the space around the vortex tube matters for flow-limiting? One guess is that it's for hot gas expansion so as to not create more pressure going into the end-port. The CFV literature talks about onset of "limiting" being a factor of the ratio of input pressure to output pressure.
You say the end-port is a "slit". What are the length x width proportions of the end port? Did you make it more slit-like with an intuition that more edge-surface would cause drag/turbulence to create a flow limiting effect?
-Dan
|
|
|
|
Post by peterberg on Jun 28, 2023 0:35:04 GMT -8
Has anyone puzzled out why the csa of the space around the vortex tube matters for flow-limiting? Yes, I did. it didn't work with a 100% csa around the tube. At the time, I tried a couple of end port sizes without much of an effect. Much later I enlarged the space and tried the end port sizes again, and to my surprise it did work. You say the end-port is a "slit". What are the length x width proportions of the end port? Measured just now, the slit is 215 mm wide and 60 mm deep, seen from the front. System size is 130 mm, in this case. The firebox and top box are both 215 mm wide. Did you make it more slit-like with an intuition that more edge-surface would cause drag/turbulence to create a flow limiting effect? Not exactly. I reasoned there would be lots of space at the sides and less in the middle so I took the easiest route and made it full width. Best performance turned out to be at 100% system csa, as said before. Just somewhat larger and it tended to thermal runaway again, somewhat smaller and it became very sluggish. One could imagine it's a tube now of system size front and rear with a fairly long wider piece in the middle. Combine that with one closed end and the firebox port at a 90 degree angle and it starts to look like a rocket engine. The only thing that's missing is the flared-out De Laval nozzle. Interesting, don't you think? |
|
